The Testing of Transformer

1
Transformer
ALSTOM Elektrik Endüstrisi A.Ş.
Barış Mah. E 5 Altı, 1801 Sok. No 104

41410 Gebze- Kocaeli / TÜRKİYE
Tel. : + 90 262 648 33 00 Fax : + 90 262 641 20 35

THE TESTING OF POWER TRANSFORMERS Page : 2
Elektrik Endüstrisi A.Ş.
PREFACE
Test procedures and equipment used for the testing and measurement of power transformers at
Gebze factory are dealt in the scope of this booklet.
The electrical characteristics and dielectric strength of the transformers are checked by means of
measurements and tests defined by standards (e.g. TS, IEC, DIN/VDE, ANSI, NEMA, BS….. etc.)
and by the requirements of customers specifications.
Summary of the tests and measurements processes are given as follows:
ROUTINE TESTS:
Pages
1. Measurement of voltage ratio and check of vector relationship 3

2. Measurement of winding resistance 6
3. Measurement of impedance voltage and load loss 8
4. Measurement of no-load loss and current 11
5. Dielectric tests 13
6. Separate-source voltage withstand test 15
7. Induced over-voltage withstand test 16
8. Partial discharge measurement 19
9. Test on on-load tap changer 22
TYPE TESTS AND SPECIAL TESTS
10. Temperature rise test 23

11. Measurement of zero-sequence impedance 27
12. Measurement of voltage and current harmonics 29
13. Measurement of insulation resistance 30
14. Measurement of capacitance and tan δ 31
15. Lightning impulse test 32
16. Switching impulse test 36
17. Measurement of acoustic sound level 38
List of tests and measuring equipment of the test field 41
Prepared by : Haluk ODOGLU December 2001 ( 4. Edition )
All rights of this document are reserved to ALSTOM Elektrik Endüstrisi A.Ş. To duplicate and to pass over to third persons without previous permission is forbidden. This document is subject to change.
ALSTOM 442/D
1. MEASUREMENT OF VOLTAGE AND CHECK OF VECTOR RELATIONSHIP
The voltage ratio (i.e. turn ratio) of the transformer is the ratio of voltages at no-load.
Purpose of the measurement: verification of no-load voltage ratios specified by the specification
and detection of any problem within the coils or tapping connections.
Measurements are carried out on all taps and on all phases.
Measuring Circuit
Turn ratio measurement is performed by two separate methods;

1. measurement by bridge method
2. measurement of voltage ratios
1. Turn ratio measurement is carried out by means of a voltage ratio measuring bridge in one-
phase basis between the winding pairs. Measurements are repeated at all phases and all
taps. During measurement only the turn ratio between the windings in which same magnetic
flux flows. In other words, the turn ratio can be measured between the winding pairs, which
are in parallel in vector diagrams (figures 1-1, 1-2, 1-3). The supply voltage is 220 V a.c.
and error of the bridge is less than ±0,1%.
~ U1
220 V U2
3
c - Transformer to be measured d - The transformer with adjustable taps

e - Zero indicator
U1 – The supply voltage of the bridge and H.V. winding (220 V, 50 Hz)
U2 – The induced voltage in L.V. winding
Figure 1-1: Measurement of voltage ratio measuring bridge
the voltage of HV winding

Theoretical turn ratio =
the voltage of LV winding
The theoretical turn ratio of the transformer is adjusted on the tapped transformer of the bridge.
‰ error indicator knob is adjusted until the balance is reached in the zero indicator. The
reading from the error indicator scale shows the difference (deviation) between the actual turn
ratio and turn ratio in ‰.
Deviation =
( measured turn ratio ) − ( theoretical turn ratio )
⋅ %100
theoretical turn ratio
ALSTOM 442/D
2. Voltage ratio measurements are generally performed by the digital instruments produced for
this purpose. In addition to voltage ratio measurement, the determination of the vector
group (connection group) and current measurements can be performed with these
instruments, it is necessary that the instrument must have 3-phase system for vector group
determination. The method of comparing the voltages of dual vectors enables the
measurement of phase shifting between the vectors.
The deviation in the turn ratio shall be < ± 0,5 %.
Vector Group
In multiphase transformers, primary and secondary connections can be either star (Y), delta (D)
or zigzag (Z), depending on the type of the transformer. The phase angle between the primary
and secondary windings changes between 0° and 360°. In vectorial denotation, when H.V.
windings shows 12 (0), the numbers of the other windings in the connection group show the
number of the clock in comparison with real or imaginary neutral point. For example, in Dyn5
connection group H.V. winding is Delta (D), L.V. winding is Star (Y) and there is a phase
displacement of 150 (5x30°) between the windings. When the vector of H.V. shows 12 (0), the
vector of L.V. shows 5 (150° lag).
The connection group is defined only for three-phase transformers. In connection group
denotation, the H.V. winding is shown first (as a reference) than the other windings are as
followed up.
The vector diagrams is also checked at the same time. The correct connection of the
measurement cables between the transformer and between the bridge verifies the vector
relationship, otherwise it is not possible to balance the bridge.
Besides the above mentioned, the check of vector relationship and the check of polarity also
could be done using a voltmeter. In this method AC or DC voltages could be applied.
The wiring connections related with the AC method are given in standards as in details. An
example to this method is illustrated in a phase diagram as below:
Example : Vector group: Dyn5 B
Measuring procedure:
a
1- three phase voltage is applied to ABC phase
2- the voltage between any two phases is measured ( e.g. AB) c
3- A and n terminals are short-circuit n
4- the voltage between B and a′ is measured a´
5- the voltage between C and c′ is measured b
A
c´ C
n´
Figure 1-2: Measuring of vector group b´
As it can be seen from the phase diagram, to obtain Dyn5 vector group the following condition
should be realised:
C c’ > AB > B a’
The other vector relationships can be checked by using the same principles.
ALSTOM 442/D
Figure 1-3: Connection symbols for three-phase transform
ALSTOM 442/D
2. MEASUREMENT OF WINDING RESISTANCE
Although, the winding resistance values are not the guaranteed values given to the customers,
they are needed in connection with the load loss measurement when the load losses are
corrected to correspond to the reference (e.g. 75°C) temperature. The resistance measurement
will also show whether the winding joints are appropriate and the windings are correctly
connected.
The winding resistances that vary with the temperature strongly, are the ohmic/d.c. resistance’s
of a winding and the resistance is computed as follows:
235 + t 2 225 + t 2
R2 = R1 ⋅ ( for copper ) R2 = R1 ⋅ ( for aluminium )
235 + t1 225 + t1
where; R2 = winding resistance at t2 temperature
R1 = winding resistance at t1 temperature
Therefore, wherever the winding resistances are stated, the temperatures during the
measurement must be given.
The resistances between all pairs of phase terminals at all tapping connections are measured.
During the measurement of the resistance, winding temperature should be correctly measured.
Direct current can be obtained from a constant-current supply or from a battery unit. The value
of the direct current should be high enough to ensure correct measurement and should be low
enough to prevent any effects on the winding temperature. In practise this value should be
greater than 1,2 x I0 and less than 0,1x In. The time constant of the measurement circuit will
depend on the ratio of L/R.
When the test object is assumed to be composed of a R resistance and L inductance which is
series connected to it, U voltage applied to this circuit will be;
R
U − t
i = (1 − e L ) where time coefficient depends on L/R ratio.
R
If the measuring current increases, it leads to the reduction of inductance due to saturation of
the core, these enables the current to reach the steady state condition in a short time.
After switching on the supply voltage to the measurement circuit, it should be waited until the
current becomes stationary, otherwise measurement errors will be occurred.
Measurement Circuit
Winding resistances can be measured by any of the following methods, by current-voltage

method or by bridge method. Measuring sensitivity can be increased by using the digital
measuring instruments. The circuit of the measurement by current-voltage method is given in
figure 2-1.
In the current-voltage measuring method, by applying the winding current through the reference
resistance in the system, the voltage drops occurred in both resistances. This voltage drop values
of reference and winding resistances are compared to determine the value of unknown
resistance (winding resistance) which can be read directly from the bridge instrument.
ALSTOM 442/D
It is necessary to care, in older avoid of very high voltages during the switching on and off the
circuit, the voltmeter shall not be kept in the circuit during this time.
a
S A
A
B b
C c
N n
V
Figure 2-1: Winding resistance measurement by Current-Voltage method
In the bridge method measurement, the principle is the comparison of unknown resistance with
a known resistance. This will be accomplished to make the current flowing through
Galvanometer to zero by bringing the arms of the bridge into the equilibrium. The lower
resistances with (<1ohm) are measured by Thomson (Kelvin) Bridge, the resistances with
relatively higher (≥1 ohm) are measured by Wheatstone Bridge. By this means, the
measurement errors are minimized.
R1 R2
R1 R2
G
G
R
R
Rx RN
R3 R4
I2
Rx rN RN R' N
A
RS A
RS
Figure 2-2: Thomson (Kelvin) Bridge Method Figure 2-3: Wheatstone Bridge
The resistance measured by Thomson (Kelvin) Bridge method.
R1
Rx = RN ⋅ ( R1 = R3 and R2 = R4 )
R2
The resistance measured by Wheatstone Bridge method.
R1
Rx = RN ⋅
R2
ALSTOM 442/D
3. MEASUREMENT OF IMPEDANCE VOLTAGE AND LOAD LOSSES
Short-circuit (load) losses and short-circuit impedance voltage are guaranteed and reported
values by the manufacturer to customers. Short-circuit impedance voltage is an important
parameter specially for the parallel operation of the transformers, where as short-circuit losses
are important from economical point of view.
This measurement is carried out to determine the load losses of the transformer and the
impedance voltage at the rated frequency and rated current. The measurements are made
separately for each winding (e.g. between 1st winding and 2nd winding for a two winding
transformer; and between 1st and 2nd, between 1st and 3rd and between 2nd and 3rd winding for
a three-winding transformer). If the tapping range is more than 5%, the measurements are
repeated on the extreme tapping also.
Measurement Circuit
a
3 A
G
3∼ B b
c
C
1
N 6 n
2 4
C
f
A A A vv
v
vv
W
W W W
1- Synchronous generator 5- Power analyser

2- Supply transformer 6- Test object
3- Current transformers for measurement C- Compensation capacitors
1- Voltage transformers for measurement
Figure 3-1: Circuit for the load-loss measurement
ALSTOM 442/D
Rated current is generally applied to the H.V. winding while L.V. winding is short-circuited. Test
current should be close to the value of IN rated current as much as possible. And the voltage
waveform should be in sinus-form with rated frequency. The voltage, current and load losses for
each phase should be measured during the test.
In case that the generator could not supply the system, the reactive power is encountered by
using the capacitor banks.
The readings have to be taken as quickly as possible to prevent the temperature changes in the
windings and the applied current should be between 25 %…… to 100 % of rated current. So
that the measurement errors are minimized.
Since the circuit is entirely inductive in large power transformers and reactors the power factor
of the system (cos ϕ) would be very small. (Cos ϕ = 0,015…0,003 or 90 - ϕ = 1°…. 10 min.),
the measurement errors arising from measurement transformers will be comparatively high.
In such cases the results of the measurements must be corrected by a correction factor.
Correction Factor
E (%)
PK1 = PK 2 ⋅ (1 − ) PK1 : corrected losses
100
PK2 : loss reading from wattmeter
E ( % ) : total error
E (%) = Eδ (%) + Ei (%) + Eu (%) Eδ ( % ) : measurement error
Ei ( % ) : current transformer turn ratio error

Eu ( % ): voltage transformer turn ratio error
 cos ϕ 
Eδ (%) = 1 −  ⋅100 where ; δ = δi - δu
 cos(ϕ − δ ) 
δi : current transformer phase displacement error
δu : voltage transformer phase displacement error
Since, phase displacement errors in measurement transformers are given in minutes.
Eδ ( % ) = + 0,0291. (δi – δu ). tg ϕ
If the measuring current is different than the “ IN ” (rated current), short circuit impedance and
load-losses are calculated referring the rated current as follows:
IN IN 2
U k = U km ⋅ Pk = Pkm ⋅ ( )
Im Im
Ukm = Measured short circuit impedance Im = Measured current
Pkm = Measured load losses Pk = Load losses at rated curent
Uk = Short-circuit impedance at rated current
ALSTOM 442/D
According to the standards, the measured value of the losses are evaluated at a reference
temperature ( e.g. 75°C ). The measuring temperature ( tm ) losses are corrected to the
reference temperature ( 75°C ) according to the standards as follows;
The d.c. losses Pdc at the measuring temperature “ tm ” are calculated using the resistance
values RHV and RLV obtained in the resistance measurement: RHV and RLV between line terminals.
DC losses = at tm measuring temperature Pdc = 1,5 ( I12 . RHV + I22 . RLV )
Additional losses = at tm measuring temperature Pac = Pkm - Pdc
The load losses at reference temperature:
t s + 75° C t +t
Pk = Pdc ⋅ + Pac ⋅ s m where ; ts : 235 °C for copper ( acc. To IEC )
t s + tm t s + 75
225 °C for aliminium ( acc. To IEC )
The short-circuit impedance:
U km
At measuring temperature ( tm ) ukm = 100 ⋅ [%]
UN
Pkm
uRM = 100 ⋅ [%] “ohmic component”, uxm = u 2 km − u 2 RM [%] “inductive component ”
SN
Pk
At reference temperature of ( 75 °C ) : uR = 100 ⋅ [%], uk = u 2 R + u 2 xm [%]
SN
Load losses and short-circuit impedance voltages measurements and corrections must be done
at rated and extreme taps.
If the short-circuit losses and voltage of a transformer measured at frequency which is different
form rated frequency, the following correction must be applied:
fN fN 2
Short-circuit impedance voltage : U k = U km ⋅ short-circuit losses : Pk = Pdc + Pac ⋅ ( )
fm fm
Where ;
Ukm : short-circuit imp. voltage at fm meas. freq. Uk : short-circuit imp. voltage at fm meas. freq.
Pac : additional losses at fm meas. freq. Pk : total short-circuit losses at fN rated freq.
ALSTOM 442/D
4. MEASUREMENT OF NO-LOAD LOSSES
The no-load test is performed when one of the windings is unconnected to power supply
(usually the H.V. winding) while the other winding is supplied with rated voltage at rated
frequency. Then the no-load losses ( P0 ) and the no-load current ( I0 ) are measured. The test is
usually carried out between 90 % - 115 % of UN voltage at equal interval and corresponding
values to the rated voltages are determined.
No-Load Losses and No-Load Current

The following losses occur at no-load
- Iron losses in the transformer core and other metal parts
- Dielectric losses in the insulation’s
- Load losses caused by the no-load current
Since the last two mentioned losses are relatively very low, they can be ignored. So that the no-
load losses are only the iron losses.
Measurement circuit :
a
G 3 A
3∼
b B
1 c C
n N
6
2 4
f
A A
A A v
vv
vv
W W W
1- Synchronous generator 4- Voltage transformers

2- Supply transformer 5- Power analyser
3- Current transformers 6- Test object
Figure 4-1: Circuit for the no-load measurement
ALSTOM 442/D
Because the losses are to be determined under standard conditions, it is necessary to apply a
wave shape correction whereby the losses are corrected to correspond to test conditions where
the supply voltage is sinusoidal. In case of non-uniform waveform, the effective value of voltage
( U ) is different than the mean value ( U’ ). There is no need for correction, if the voltmeter
readings are the same for all.
The test voltage wave shape is satisfactory if the difference between readings U’ and U are
less than 3 %.
The test voltage ( U’ ) is adjusted by the “mean value” voltmeter. Then the hysterisis losses can
be measured correctly. But the Eddy-current losses should be corrected.
Pm = Po ⋅ ( P1 + k ⋅ P2 ) Pm : measured losses
P0 : no-load losses at sinusoidal voltage
2
U 
k = ' P1 : ratio ( as a percentage ) of hysterisis losses to total iron losses
U 
P2 : ratio ( as a percentage ) of Eddy-current losses to total iron losses
It is assumed that for cold oriented steel sheets at 50 and 60 Hz , P1=P2= 50 % This yields
Pm
P0 = where ; P1 = P2 = 0,5
P1 + k ⋅ P2
U ′ − U 
According to IEC 60076-1; Pm = P0 . (1+d) where; d =
 U ′ 
During the no-load loss measurement, the r.m.s value of no-load current is also measured at
the same time. For a three-phase transformer, the average of the three phase currents is taken.
During the measurements the connections of the transformer shall be as foreseen in the service.
The transformer could be magnetised by d.c. before performing the no-load test. Therefore, the
transformer shall be supplied with a higher voltage during a suitable duration (a few minutes)
for demagnetisation of the core then the measurement shall be made.
Since the no-load currents are not symmetrical and generally not equal in magnitude. They
have also different phase angles in three phases, so that the indications on the wattmeters will
not be equal. The indication on one wattmeter can be zero ( 0 ) or even reach to ( - ) negative
value.
If the measurement are being made on delta connected windings, one of the current can be
absolutely greater and the other two can be smaller and their magnitude approximately equal
to each other. On star connected windings, the measured current on the middle phase can be
smaller and the outer phase currents can be greater.
When the analysis is made on phase currents, due to non-linear and dissymmetrical structure of
the core, dissymmetrical distortion will be also detected on the currents, which are happened as
current harmonics. ( Please see the harmonic measurement section ). The measurements are
made normally at room temperatures.
ALSTOM 442/D
5. DIELECTRIC TESTS
The standard dielectric requirements are verified by dielectric tests. They shall, where applicable
and otherwise agreed upon be performed in the sequence as given below:
- switching impulse test: the test is intended to verify the switching impulse withstand
strength of the line terminals and its connected winding(s) to earth and other windings and
along the windings.
- lightning impulse test: the test is intended to verify the impulse withstand strength of the
transformer, when the impulse is applied to its line terminals.
- separate source withstands voltage test: the test is intended to verify the AC
withstands strength of the line and neutral terminals and their connected windings to earth
and other windings.
- induced AC withstand voltage test (short ACSD and long duration ACLD): the test
is intended to verify the AC withstand strength of the each line terminal and its connected
winding(s) to earth and other windings, the withstand strength between phases and along
the winding(s) under test.
- partial-discharge measurement: it verifies partial-discharges free operation of the
phases and along the winding(s) under test.
Transformer windings are identified by their maximum operating voltage Um and their
corresponding insulation levels. Insulation levels and applicable dielectric tests are given in the
table below.
Category Highest voltage Tests

of for equipment
winding Lightning Switching impulse Long duration Short duration Separate
Um kV impulse AC AC source
( LI ) ( SI ) ( ACLD ) ( ACSD ) AC
Uniform Um ≤ 72,5 type not applicable not applicable routine routine

insulation ( note 1 ) ( note 1 )
72,5 <Um ≤ 170 routine not applicable special routine routine
Uniform
and non- 170 < Um < 300 routine routine routine special routine
uniform ( note 2 ) ( note 2 )
≥ 300 routine routine routine special routine
Note 1 : In some country, for transformer with Um ≤ 72,5 kV , LI tests are required as routine tests, and ACLD
tests are required as routine or type tests.
Not 2 : If the ACSD test is specified, the SI test is not required.
The transformers which have one or more non-uniformly insulated windings, the test voltages
for induced voltage test, and for the switching impulse test if used, are determined by the
winding with the highest Um value, and the windings with lower Um values may not receive
their appropriate test voltages. This discrepancy should normally be accepted. If the ratio
ALSTOM 442/D
between the windings is variable by tappings, this should be used to bring the test voltage for
the winding with lower Um voltage as close as possible to the appropriate value.
The details of the partial discharge are given in section 8.
If LI is requested to include the chopped wave in this case the peak value of the chopped
impulse shall be 10 % higher than for the full impulse ( i.e 10 % more ).
For transformers with a high-voltage winding having Um>72,5 kV, lightning impulse tests are
routine tests for all windings of the transformer.
Repeated dielectric tests:
For transformers which have already been in service and have been refurbished or serviced,
dielectric tests shall be repeated at test levels of 80 % of the original values, unless otherwise
agreed upon, and provided that the internal insulation has not been modified. Long duration
AC induced test ( ACLD ) shall always be repeated at 100 % test level. Repetition of tests
required to prove that new transformer, having been factory tested, is always performed at
100% of test level.
ALSTOM 442/D
6. SEPARATE-SOURCE VOLTAGE WITHSTAND TEST

The purpose of the test is testing the insulation between the windings and the insulation between
windings and earthed parts (tank, press iron, etc.), under the temporary and switching over-
voltages conditions which may be occurred during the service.
Test Circuit
8
5
1 4
A a
Vef V B b
∼
C c
2
N n
A V 3 6 7
1- Transformer with adjustable voltages 5- Capacitive voltage divider

2- Current transformer and Ammeter 6- Voltmeter (r.m.s. value)
3- Input voltage voltmeter of test transformer 7- Voltmeter (Peak value/√2)
4- Test transformer 8- Transformer under test
Figure 6-1: Test circuit for separate-source voltage withstands test
The tests, which are made with separate source voltage, shall be made at rated frequency or at
a frequency not less than the 80 percent of the rated frequency. This allows the testing of 60 Hz
transformers at 50 Hz. The waveform of voltage must be sinusoidal as much as possible and
must be single phase.
The voltage is measured using a capacitive voltage divider in conjunction with voltmeter
response to peak values. The peak voltmeter indicates the peak value divided by √2. The test
duration is one minute. The test voltage is applied to the winding which is going to be tested
(all terminals of this winding is short-circuited) while the terminals of the other windings are
connected to each other, the windings that are not under test voltage, tank and core also should
be grounded. Secondary winding terminals of the bushing type CT’s shall be connected to each
other’s and grounded in order to prevent unwanted sparkings.
The current should remain constant during the test.
The test is successful if no collapse of the test voltage occurs.
This test is applied to the uniformly insulated windings and to the star point (neutral point) of the
graded insulated (non-uniform) windings. Every point of the windings to which the voltage
applied; is tested with the test voltage.
The line terminals of non-uniform insulated windings are tested by induced overvoltage
withstand test (section 7).
ALSTOM 442/D
7. INDUCED OVERVOLTAGE WITHSTAND TEST
The purpose of the test is testing the insulation between the phase windings, turns, coils, tapping
leads and terminals, for non-uniformly insulated windings also the insulation between these
parts and earth, under the temporary and switching over-voltages conditions to which the
transformer may be subjected during its life time.
Normally, the excitation voltage is applied to the terminals of the low voltage winding while the
terminals of the other windings are left open or grounded in one point.
Since the test voltage is much higher than the rated voltage, the frequency of the test voltage is
chosen at least two times greater than the rated frequency without causing the over saturation in
the core.
The test voltage is measured with capacitive voltage divider connected to H.V. terminal or the
test voltage can be read from the voltmeter through voltage transformer in the low voltage side,
which is calibrated with the voltage divider. Another method is to measure the test voltage from
the peak value voltmeter, which is connected to the test tap of H.V. condenser bushings. The
peak voltmeter indicates the peak value divided by √2.
The test duration not being less than 15 seconds is determined by the following formula;
Rated frequncy
120 second x
Test frequency
If no flashover voltage collapse and abnormal increase in the current occurs during the test,
then the test said to be satisfactory.
Short duration induced AC voltage test ( ACSD ) :
a) Uniformly insulated windings
Test connection is essentially the same as in service, a three phase symmetrical voltage is
supplied. The test voltage is generally twice as the rated voltage. However, the voltage
between line terminals of any winding shall not exceed the rated short duration power
frequency withstand voltage. The tappings of the transformer should be accordingly.
The voltage is measured from terminals to earth or between terminals of the low voltage
winding using precision voltage transformers.
Test circuit W C
c
G V b B
3~
1- synchronous supply gen. U a A
2- test transformer 1
3- current transformer and ammeter
4- voltage transformer and voltmeter 5 N
5- transformer under test 2 N A
V
3
4
Figure 7-1: Test circuit for induced over-voltage withstands test on uniformly insulated winding
ALSTOM 442/D
Transformers with Um < 72,5 kV normally , no partial discharge measurements

are performed during this test. Test duration and voltage are explained above.
Transformers with Um > 72,5 kV this transformers shall all, if not otherwise agreed,
be tested with partial discharge measurement. The partial discharge performance shall be
controlled according to the time sequence for the application of the voltage as shown in
figure 7.2 :
U 2 = 1,3 ⋅ U m / 3 phase – earth and U2 = 1,3 . Um phase - phase
A : 5 min C
B : 5 min
C : test duration D
D : ≥5 min B
E : 5 min
A E
Um : Highest voltage for
equipment U2 U1
U2
1,1 ⋅ U m / 3 1,1 ⋅ U m / 3
time
Figure 7.2: Test duration voltage – time curve
b) non-uniformly insulated windings :
For three phase transformers, two test sets are required ;
1. A phase-to-earth test with rated withstand voltages between phase and earth with
partial discharge measurement.
2. A phase-to-phase test with earthed neutral and with rated withstand voltages
between phases with partial discharge measurement. The test shall be carried out
in accordance with uniformly windings ( subclause a ).
On single –phase transformers, only a phase-to-earth test is required.
The test sequence for three phase transformer consist of three single-phase applications of
test voltage to the individual phases with different points of the winding connected to earth
at each time.
At this type of windings, induced overvoltage test and separate source voltage withstand
test ( at the pase terminals ) are performed in the same time.
For the three single-phase tests for the phase-to-earth insulation ; U 2 = 1,5 ⋅U m / 3
For the partial discharge performance evaluation, during the phase-to-phase test,
measurements should be taken at U2 = 1,3 . Um . For Um = 420 kV and 550 kV
transformers with test values of 460 kV and 510 kV, the PD evaluation level should be
reduced to U2 = 1,2 . Um during the phase-to-phase test and U 2 = 1,2 ⋅ U m / 3 during
the phase-to-earth test.
ALSTOM 442/D
Test circuit W C
c
G V b B
3~
1- synchronous supply gen. U a A
2- test transformer 1
3- current transformer and ammeter
4- voltage transformer and voltmeter 5 N
5- transformer under test 2 N A
V
6- Capacitive voltage divider 3
4
V 6
Figure 7.3: Test circuit for induced over-voltage withstands test on non-
uniformly insulated winding of a three-phase transformer
The test circuit given in fig. 7.3 is for the transformer which HV neutral point is insulated
according to 1/3 of the test voltage.
Long duration induced AC voltage test ( ACLD ) :
Uniformly and non uniformly insulated windings.

Three- phase transformer shall be tested either phase-to-phase in a single-phase connection or
in a symetrical three-phase connection.
The neutral terminal ( if present ) of the winding under test shall be earthed. The other separate
windings, if they are star-connected they shall be earthed at the neutral, and if they are delta-
connected they shall be earthed at one of the terminals or earthed through the neutral of the
supplying voltage source. The duration and the voltage levels are given in fig. 7-4.
A : 5 min C
B : 5 min D
C: test duration B
D: for Um>300 kV 60 min
for Um<300 kV 30 min A E
E : 5 min
U2 U1 U2
Um : Highest voltage for
equipment 1,1 ⋅ U m / 3 1,1 ⋅ U m / 3
time
Şekil 7.4: Long duration induced overvoltage test voltage – time curve
During the whole application of the test voltage, partial discharges shall be measured. The
details of the PD measurement are explained in section 8.
The voltages to earth shall be ; U1 = 1,7 ⋅ U m / 3 and U 2 = 1,5 ⋅ U m / 3
The partial discharge measurements shall be performed at all HV terminals.
The details of the evaluation of the tests and partial discharge measurements are given in
standards ( e.g IEC 60076 – 3 )
ALSTOM 442/D
8. PARTIAL DISCHARGE MEASUREMENT
The purpose of this test is to measure the partial discharges in the test object produced by the
application of AC voltages during the tests. This test gives a comprehensive information about
the quality of the insulating materials and the design.
The followings are determined during the measurement of partial discharges;
- To determine the existence of a definite partial discharge in the test object at a

predetermined voltage.
- By increasing the applied voltage at where the partial discharging begins (partial discharge
inception voltage) and by decreasing the applied voltage at where the partial discharging
extinguishes (partial-discharge extinction voltage).
- To determine the magnitude of the partial discharge at a predetermined voltage.
The mentioned partial discharges (which do not cause flashover between the electrodes) are the
discharges in a certain area of the insulation between the conductors of the test object. These
discharges may occur in the gaps of the insulating enviroment, in the gaps of the solid-
materials or in the contact surfaces of two different insulations. This discharge can be captured
as a single current impulse in the outer region. Although these discharges do not cause
permanent deteriorations in the insulating media since their energy is relatively small, the
thermal energy of the discharges shall cause depreciation, aging and deterioration in the
insulating media.
The electrical discharge magnitude at the partial discharge point is not a direct measurement
for deterioration of the insulating material in this region. Besides the numerical value, the
intensity and the waveform of the impulse, regional discharge concentration, the manufacturing
and the placing of the insulation also effects the situation.
The above evaluations prove that the “partial discharge” measurement is a implementary
method to check the quality of whole insulation.
Creation of partial discharge and measured magnitudes
In a insulation arrangement, an analogue schematic drawing about the partial discharge takes
place in a certain region is shown below.
As it can be seen in the figure 8.1, the impulse occurred in the discharge point produces ∆U
voltage across the line terminals. This yields a measurable “q” charge in the measurement
impedance.
This is called apparent charge and is given as pC (Pico-Coulomb). This apparent charge is not
equal to the charge in the fault region but it is proportional, however this proportion could not
be fixed yet.
In the measurements; ∆U voltage drop, the average value of the partial-discharge current, the
partial-discharge power, number of impulses in a time interval, and the inception and extinction
voltages of the partial discharge can also be determined.
ALSTOM 442/D
Z Z
Discharge region
C1
U a) G
U C
insulation R
C2
a)
b)
U: applied voltage Z: supply circuit impedance

R : discharge resistance G : discharge gap
C1: capacitance of discharge region C : Capacitance of the other parts of the insulation
C2: capacitance of the insulation series connected with the discharge region
Figure 8.1 a) An insulation having gas-filled cavity b) equivalent circuit
Measurement principle and circuit

Partial discharge measurement principles in a transformer and a produced circuit, according to
IEC-Publication 60270, are explained below.
qo
a
A
G b ZZmm
∼ B
c ZZmm
C
c ZZmm
5 e N h
d
f Ê
osc. i
V j
1- supply generator 6- measuring impedance
2- supply transformer 7- selector switch
3- test transformer ( test object ) 8- measuring equipment and oscilloscope
4- voltage transformer and meas. Circuit qo- calibration generator
5- filter
Figure 8.2: Connection circuit for partial discharge measurement
ALSTOM 442/D
The circuit that is given at figure 8.2 is a very useful method called BUSHING TAP.
The calibration of the circuit has to be made before starting the measuring. For this reason first
of all a calibration generator is needed. Calibration generator is producing a q0 charge, with a
definite value. It is connected parallel to the test object. The q0 charge, which is produced by
calibrator, is observed from the measuring equipment. This operationmust be repeated for all
terminals of the transformer.
qo K : correction factor
K=
qom q0 : the charge produced by the calibrator
q0m : the charge observed at measuring equipment
P.S.: It must be noted that the transformer must be de-energized during the calibration process.
After the transformer has been energised, the value of partial discharge that was observed from
the measuring equipment, is multiplied by the correction factor ( K ) of every terminal and the
apparent partial discharge value is found.
K : correction factor
q = qm . K
qm : the charge observed at measuring equipment
q : the real apparent charge
The wide-band or narrow-band measuring equipment’s are used for partial discharge
measurement. Wide-band consists of the frequency spectrum from 40 kHz up to several 100
kHz, and narrow-band frequency spectrum is between 9÷10 kHz.
i.e.: 0,2 MHz, 0,5 MHz, 1,9 MHz.
The wide-band measuring equipments are to much effected from environmental influences
where as narrow-band ones is effected considerably little.
The partial discharge measurements of the transformer is made during the induced voltage test.
Test and measuring circuit, test duration and voltage levels, and evaluation criterions are given
section 7, induced overvoltage test.
ALSTOM 442/D
9. OPERATION TESTS ON ON-LOAD TAP CHANGER
After the tap-changer is fully assembled on the transformer, the following tests are performed at
(with the exception of clause b) 100% of the rated auxiliary supply voltage;
a) 8 complete operating cycles with the transformer not energised.
b) 1 complete operating cycle with the transformer not energised, with 85% of the rated
auxiliary supply voltage.
c) 1 complete operating cycle with the transformer energised at rated voltage and
frequency at no-load.
d) 10 times tap-changer operations with ±2 steps on either side of the principal tappings
with as for as possible, the rated current of the transformer, with one winding short-
circuited.
ALSTOM 442/D
10. TEMPERATURE RISE TEST
The purpose is, to check that the temperature rises of the oil and windings do not exceed the
limits agreed on or specified by the standards.
The supply and measuring facilities as well as the measuring circuits are the same as in the
load loss measurement (section 3) and in the measurement of windings resistance (section 2).
Simplified temperature distribution diagram is given in figure 10-1.
θo
θg
∆θ o
θo
θhs
R θg
θ wmax
① ② θ oavg
θw
θc
∆θ oavg ∆θ wo
∆θ w
θa θ
Figure 10.1: A simplified temperature distribution for a transformer
θo = maximum oil temperature ( under cover )

∆θ o = maximum oil temperature rise ∆θ o = θ o − θ a
θa = ambient temperature
θw = the average winding temperature
∆θ w = the average winding temperature rise ∆θ w = θ w − θ a =∆ θwo + ∆ θoavg
θg = temperature of oil going into the cooler
θc = temperature of oil coming from the cooler
θ w max = maximum winding temperature
R = cooler
θ oavg = the average oil temperature rise
∆θ wo = the temperature difference between winding and oil
∆θ oavg = the average oil temperature
θhs = hot-spot temperature
ALSTOM 442/D
a) The measurement procedure
The necessary precautions should be taken for the effects ( i.e. warm or cool air circulation )
which would affect the transformer under temperature-rise test.
The current and voltage measuring principles, during this test, are the same as in the load loss
measurement ( section 3 ). The temperature rise test will, unless otherwise specified, be carried
out on the maximum current tapping.
Since, it is necessary to record the temperature rises and the ambient temperature through the
test, thermometers are placed in a thermometer pocket in the cover, in and going out of the
cooler and 1 to 2 meter away from the transformers. The temperatures in these thermometers
are measured and recorded when the transformer is in cold position before starting the test.
Before starting the test, the winding temperature ( cold resistance measurement ) is measured
and recorded.
The transformer is supplied with a voltage and current which constitutes the sum of short-circuit
losses at the maximum loss tap and no-load losses in order to achieve to the service conditions.
For a multi-winding transformer, the temperature rise requirement refer to rated power in all
windings simultaneously if the rated power of one winding is equal to the sum of the rated
powers of other windings.
In certain cases, if it suits, first part of the test can be a few hours shortened by switching of the
cooling system.
The maximum values of the current and the voltage during the supply are as follows;
Po + Pk Po + Pk
Supply current Id = I N ⋅ Supply voltage Ud = Uk ⋅
Pk Pk
in which;
IN = Rated current (the current of the tap in which the test is performed)
Po = No-load losses , Pk = Load losses
The test is performed separately in two parts:

a) Total loss injection ( 1. Part of the test ) : Supplied with total losses. The test is continued
until a steady-state oil temperature rise is established ( i.e. the difference between top oil
temperature and ambient temperature is less than 1° C for one hour ). This period is
called as first part of the test. The supply values and the temperatures of different points
are recorded at suitable time intervals.
b) Rated current injection ( 2. Part of the test ) : When the top oil temperature rise has been
established, the test shall immediately continue with rated current supply one hour. This
period is called as second part of the test. The supply values and the temperatures are
recorded as above. At the end of one hour, supply is disconnected and the hot-resistance
of windings are measured. The test connection is changed for carrying out the resistance
measurement and after the inductive effects have disappeared the resistance time-curves
are measured for a suitable period of time than by extrapolation method the resistance
value of the winding at the instant of switching off the supply is determined.
After disconnection of the test current, the pump circulation and fan ventilation are continued.
ALSTOM 442/D
b) Measurement of ambient temperature (Temperature of cooling air or water)
The cooling ambient temperature is measured by means of at least three thermometers or

thermocouples. Measurement is done by placing the thermometer or thermocouple in a
container filled with oil which has 2 hour time-constant. Containers should be protected
against any air circulation and thermic ray. They should be replaced 1 and 2 m away from the
three sides of the transformer at a height of half way up the transformer coolers.
For forced-air cooled transformers the temperature of the in going air is measured. If water is
used as cooling medium, the water temperature at the intake of the cooler ( in the thermometer
pocket ) is the reference temperature.
The values of cooling ambient temperature ( cooling air or water temperature ) taken at every
1/2 hour in the last quarter of the test are used for temperature rise calculations.
c) Determination of the temperature rise of oil
The top oil temperature is measured by a thermometer placed in an oil filled thermometer
pocket on the cover. The difference between the max. measured temperature and ambient
temperature is : θ o
1
θ avg = θ o − ⋅ (θ g − θ c ) the average oil temperature
2
∆t =θ −θ the average oil temperature rise
oavg oavg a
The temperature of oil coming in and going out of the cooler is measured by means of
thermometers which are fitted to the pipes of the cooler. When the transformer has separate
cooler, oil inlet and outlet temperatures are measured on the inlet and outlet pipes near to the
transformer tank.
If the test object during the test can not be supplied by the current which encounter the total
losses of the insufficient power, in this case test losses ( ensuring that not less than the 80
percent of the total losses ) are computed as follows;
X ∆θ0N : total-losses temperature rise at PN ( rated value )

P 
  ∆θ0M : test-losses temperature rise at PM ( test value )
∆θ oN = ∆θ ⋅ N
oM  P 
 X : for distribution tr. ( natural cooling, rated power < 2500 kVA ) = 0,8
 M 
ON-cooling = 0,9 OF and Water cooling = 1,0
d) Determination of the temperature rise of the winding
The transformer under test is continuously supplied by IN rated current after the oil temperature
is reached to the equilibrium. This is necessary to adapt the equilibrium between the winding
and the oil into the conditions when the transformer is in service. At the end of this time, the
circuit is switched off and the winding cooling curve is obtained by the method of winding
resistance measurement.
The winding temperature rise is calculated as follows;
ALSTOM 442/D
θ2 : Winding temperature at the instant of switching off

R
θ 2 = 2 (235 + θ1 ) − 235 where θ1: Winding ( cold ) temperature at the beginning of the test
R1 ( avarerage oil temperature )
R2 : the resistance at θ2 temperature
R1 : the resistance at θ1 temperature
P.S.: For aluminum windings 225 K should be taken instead of 235 K.
When the supply of IN rated current for 1 hour is the second part of the test, the temperature of
oil decreases during this time. The relation between the temperature of the winding and the
average temperature of the oil shall be calculated with this decreased temperature.
Then: ∆θω0 : the temperature differance between winding and oil

∆ θwo= θ2 − θoavg(IN) where; θ2 : winding temperature
θ0avg(IN) : the average oil temperature at the end of 1 hour IN supply
The temperature θavg(IN) : oil temperature rise is determined as in clause c)
The difference between the winding temperature and the average oil temperature
in the second part of the test is as follows ;
θ w the average winding temperature is calculated when the ∆θ wo , is added with the ∆θ oavg
from the clause c ) in the first part of the test (the supply of transformer with Id ).
∆ θw = ∆ θwo + ∆ θoavg
If the IN , rated current, can not be reached because of insufficient supply system, the difference
between the temperature of the winding and the temperature of the oil is corrected as follows:
y ∆θω0Ν : the temperature differance at the IN rated current

I 
∆θ woN = ∆θ woM ⋅  N 
 where; ∆θω0Μ : the temperature differance at the test current IM
 IM  y .. exponent : ON and OF cooling = 1,6
OD cooling = 2,0
The maximum temperature occuring in any part of the winding insulation system is called the
“hot-spot temperature”. This parameter represent the thermal limitation of loading of the
transformer. The winding hot-spot temperature rise versus ambient is computed as follows:
θhs = θoavg + Chf ⋅ ∆wo Chf : hot-spot factor
Hot-spot factor is 1,1 in distribution transformers and 1,3 in medium size power transformers.
When the transformer is supplied with total losses in the first part of the test, there is no need to
make any correction if the test frequency is different than the rated frequency, but for the
loading with rated current for one hour in the second part of the test, the correction must be
made in accordance with the below equation;
2 IM : test current IN : rated current

f 
Pdc +  N
f M 
. Pac fM : test frequency fN : rated frequency
IM = I N .

Pdc : dc losses Pac : additional losses
Pdc + Pac
ALSTOM 442/D
11. MEASUREMENT OF ZERO-SEQUENCE IMPEDANCE
The zero-sequence impedance is usually measured for all star or zig-zag -connected windings
of the transformer. The measurement is carried out by supplying with rated frequency between
the parallel connected phase terminals and the neutral terminal. The zero-sequence impedance
per phase is three times the impedance measured in this way.
Uo
Zo = 3 ⋅ Ω / Phase
I
A a
B b
G
∼ C c
n
N
A V
Figure 11.1: Circuit for zero-sequence impedance measurement
Zero-sequence impedance is related with connection group and manufacturing properties.

Zero-sequence impedance consists of two parts as Ro ( real ) and Xo ( imaginary ), due to
Ro<< Xo, Ro can be neglected. This yields the zero sequence impedance equals to zero-
sequence reactance.
Zero-Sequence impedance, only winding which star point is brought out can be measured.
Measurement is done at principal tapping when the active part of the transformer is installed in
the tank.
The value of zero-sequence impedance is infinity for the windings which are delta connected or
the star point is not brought out.
When the other winding of the transformer is delta connected or a delta connected tertiary
winding exists, in the measurement of zero-sequence impedance, the star point of the
transformer can be loaded up to max. rated current. At this instant, the Uo test voltage will be
between 15 % and 27 % of the phase-neutral voltage of the transformer. In cases where
counter magnetic flux does not exists e.g. in star -connected transformers which do not have
tertiary windings, this test current shall be max. 0,3 x IN to avoid excessive temperature of
metallic constructional parts.
It can be said that for the transformers which both windings are star connected and which star
points are brought out, there are two different zero-sequence impedance’s.
ALSTOM 442/D
a- No-load zero-sequence impedance Z 00

When one of the star-connected windings is measured, the terminals of the other winding is
left open.
b- Short-circuit zero sequence impedance Z ok

When one of the star-connected windings is measured, the terminals of the other winding
and the star point terminals are short circuited.
Zero-sequence impedance may be given as a percentage of the rated phase impedance. In this
case:
IN
zo = Z o ⋅
UN
zo = relative zero-sequence impedance (%)

IN = rated phase current (A)
UN = rated phase-neutral voltage (V)
Zo = Zero-sequence impedance (Ω/ phase)
ALSTOM 442/D
12. MEASUREMENT OF THE CURRENT AND VOLTAGE HARMONICS
Generally at the rated current the ratio of the current harmonics to the rated current is less than
1%. Therefore; this is not so important for the transformer operators, but this measurement can
be required to have an idea about the value.
The magnitude of harmonic component, depends on the core material, the degree of the
excitation, the design of the core, the connections of the windings and the impedance of the
supply circuit.
The measurement of current and voltage harmonics is made during the measurement of no-
load losses and no-load current (section 4). Circuit diagram of harmonics measurement is
given in fig. 12.1.
The supply voltage waveform in the test laboratory shall be exactly sinusoidal. The supply
voltage form may differ from sinusoidal form due the deterioration’s in the no-load current. In
order to overcome this event, the magnetic characteristics of the test generator and transformer
stay in the linear region, by choosing the appropriate connection-groups. Since the measuring
voltage and measuring current are taken through voltage and current transformers to the
analyzer, the operation regions of the measuring transformers shall be linear not to produce
additional harmonics.
Measurements are made for each phase. Usually 3 rd, 5 th, 7 th and 9 th harmonics which
are comparatively stronger, are measured.
n
The effective value of the no-load current is: I eff = ∑ Ii 2
i =1
G
G 1
Test.
~
33∼
object
A A A 4
1- Current transformer 2
Analyzer
2- Voltage transformer
3- Analyzer
4- Test object 3
V
Figure 12.1: Circuit diagram of harmonic measurement
ALSTOM 442/D
13. MEASUREMENT OF INSULATION RESISTANCE
The purpose of the measurement is to determine the leakage current of the insulation
resistance. This current is changing with the moisture, impurity contents and temperature of the
insulation.
Beside the result of the measurements, the comparison of the periodical measurement give the
information about the condition of the insulation. In order to compare they must be at the same
temperature ( for example at 20°C reference temperature ).
“Time resistance method” in the insulation resistance measurement is one of the best methods,
that is simple and gives correct results.
The insulation resistance is measured by means of an insulation resistance meter which apply a
voltage 1000 V dc or 5000 V dc. Each winding is measured separately by connecting the
voltage between the winding to be tested and earth. While the other windings are connected to
the guard circuit of the test instrument. The temperature and humidity are recorded during the
test.
The Resistance values R15 , R30, R45 and R60 are taken at 15 s , 30 s, 45 s and 60 s after apply the
voltage. Furthermore, the ratio of the insulation resistance R60 to the insulation resistance R15
(or R30 ) is stated as absorption ratio in the test report.
Readings are referred to 20 °C by multiplying the value at transformer oil temperature θ

(ambient ) by correction factor given in table below:
θ Correction θ Correction
°C factor °C factor
-10 0,13 35 2,80
-5 0,18 40 3,95
0 0,25 45 5,60
5 0,36 50 7,85
10 0,50 55 11,20
15 0,75 60 15,85
20 1,0 65 22,40
25 1,40 70 31,75
30 1,98 75 44,70
ALSTOM 442/D
14. CAPACITANCE AND TAN δ MEASUREMENT “INSULATION POWER FACTOR TEST”
All insulating materials used in practice have slightly small dielectric losses at rated voltage and
at rated frequency. These losses are fairly low for good insulating materials. This loss changes
proportionally with the square of the applied voltage. Insulation, in terms of basic circuits
elements, is shown in figure 14.1.
I
V1 = U
I
C I
1 Ic
U R=
IR G
IC δ
V
V2 = 0
IR
a) b) c)
Figure 14.1 a) Insulation b) Equivalent circuit c) Vector diagram
As it can be seen from figure 14.1 the angle δ between the total current “ I “ and capacitive
current “ Ic “ is a basic value.
Insulation angle is dependent on the thickness, the surface and the properties of the insulation
material ( the pores, impurities and humidity which cause the ionization in the insulation
material ).
Generally, the conditions and the reasons will cause a decrease in the dielectric withstand of the
insulation. For this reason, the power factor measurement of the insulation at certain frequency
gives a basic idea about the insulation. The measurements to be made during the service are
one of the most important indications, showing the ageing of the insulation and the
contimination of the oil.
The active losses of the circuit is as follows;

2
P= U⋅I⋅Cosϕ = U . C. ω. tan δ
Capacitance, tan δ, active losses and Cos ϕ can be measured by bridge methods or power
factor ( Cos ϕ ) measuring instrument at definite voltages.
Measurement is performed between the windings and the tank, and the test temperature is
recorded; then according to desired reference value the necessary corrections are done.
ALSTOM 442/D
15. LIGHTNING IMPULSE TEST
The purpose of the impulse voltage test is to confirm that the transformer insulation’s withstand
the lightning overvoltages which may occur in service.
The power transformers used in high voltage systems at any time may be affected by the
atmospheric discharges. The magnitudes of the lightning over voltages always depend on the
impulse current and the impulse impedance where the lightning impulse occurs. This value is
several times of operating voltage.
Impulse voltage is produced by a “ impulse voltage generator ” in the laboratory. For oil type
transformers, this impulse voltage is stated as with (-) negative polarity and the waveform at the
line terminal shall be T front / Ttail = = 1,2 ± 30% / 50 ± 20% µS. Besides the full wave form (figure
15.1) in chopped wave at the tail, the chopping time shall be between 2...6 µS. (figure 15.2).
T1 = 1,2 m 30% µS
T2 = 50 m 20% µS
Figure 15.1: Full wave lightning impulse
Tc = 2....6µS
Figure 15.2: Chopped wave at the tail
ALSTOM 442/D
Lightning impulse voltages are applied to the line terminals successively. The number and
application method of the lightning impulse voltages are stated in the standards. The other line
terminals and the neutral line terminal shall be grounded directly or through a small resistance
(fig. 15.3 and 15.4).
In three-phase transformers, unless it has been agrred on a particular tapping, impulse test are
performed at he two extreme tappings and the principal tapping. Each phase tests are
performed at min., max. and principal tapping.
Sometimes in the LV windings of the higher rated transformers, it might not be possible to reach
half time-value on tail as it is defined in the standards. In such cases, suitable resistances may
be connected between the windings which are not under the test and the ground. The
resistance’s must be so selected, according to IEC 76-3, that the voltage of these terminals of
the windings against the ground must not be greater than the 75 % of the test voltages of the
winding terminals and the value of the resistance must be maximum 500 Ω
Rs
A B C
Cs
Rp
osc
a b c
osc
R R osc
Impulse Generator Test Object Measurement Circuit
Figure 15.3 : Connection diagram for the lightning impulse test
Some connection diagrams used for lightning impulse test are given in figure 15-4.
1U 2U 1.1 2.1 1.1 1U 1V 1W

2.1
RW RW
1V 2V 2V 2
OSC
R 1N
Rm Rm Rm
OSC
OSC Rm
Rm
ALSTOM 442/D
1N RW RW RW
1.1
1U 1V 1W
2.
Rm 1U 1V
1W
1N OSC
Rm Rm
1V
RW1
1U 1V 1W
1U 1W
1U 1W OSC
Rm
Rm
Rm
Rm RW2 RW2 RW2
Figure 15-4 : One and three phase transformer impulse test
The voltage dividers used for measuring of impulse voltages generally are made of three types.
1- Ohmic voltage divider

2- Capacitive voltage divider
3- Mixed (ohmic capacitive) voltage divider
Although the types are changing according to the aim and place of using the most widely used
voltage divider is R damped capacitive voltage divider.
To measure the impulse currents, non-inductive, ohmic resistances are used. Usually its value
varies from 0,1 Ω to 20 Ω.
Coaxial cables are used to transmit the measurement signals to peak-value voltmeter and to
the oscilloscope.
If the application of chopped-wave is foreseen; a system which chopped the wave is added to
the impulse circuit. In modern impulse voltage circuits, this is usually a “ Multiple chopping
device ”.
ALSTOM 442/D
In beginning impulse voltage test is started with a value of 50 percent of the test voltage which
wave form is determined by the oscillograme. After obtaining the form which is acceptable
according to the the standards, first a “ Reference Impulse ” which has the magnitude of 50%
percent of the test voltage, is applied, then “ Full Impulse ” which has a value of 100% are
applied at certain times which the standards are stated.
The magnitudes of applied voltages are determined numerically on the peak-value voltmeter
through a voltage divider. Besides, the time-change of the applied voltage, and the changes of
the current leaking from the winding under test to the earth or the capacitive current leaking
from the windings which are under test, to the earth; are detected with photographs by means
of an oscilloscope.
For the evaluation of the impulse voltage tests, the evaluation of oscillographic records is a
most widely and most used method that is stated in the standard. For evaluation purposes, the
oscillographic record of the reference waveform which has small magnitude ( 50% ......75% )
should completely coincide with the oscillographic record of the wave form which has full
magnitude ( 100% ).
In some cases, there might be inconsistencies in the oscillogrammes due to the effects caused
by the arrangement of the test circuit, external disturbances and/or earthing circuit. Such
deviation should not be considered as indication of a failure.
Differences in the instant of firing of the stages in the impulse generator may give rise to initial
high frequency oscillations in the front part of the voltage vaweform.
Small differences in the wave pattern because of the chopping time may cause deviations after
chopping. These should not be considered as symptom for any failure.
If the impulse voltage to the neutral points is requested in technical specifications; the method
of impulse voltage application on this point are stated by two ways in the standards.
a) The application of a voltage, which will produce a determined impulse voltage magnitude in
the neutral point, to the parallel connected line terminals.
b) The direct application of determined impulse voltage to the neutral point.
When an impulse is applied to the neutral terminal “stated in b”, a voltage waveform which has
longer front duration ( up to 13 µS ) is permissible according to the standards.
ALSTOM 442/D
16. SWITCHING IMPULSE TEST
The purpose of the switching impulse test is to secure that the insulation between windings,
between windings and earth, between line terminals and earth and between different terminals
withstand the switching overvoltages which may occur in service.
The switching impulse voltage is simply produced by conventional impulse voltage generator.
The polarity of the voltage is negative, and form of the voltage shall be T1 /Td / T2 ≥
200/200/500 µS according to IEC 76.3 (fig. 16.2).
Test conditions determine the choice of transformers tapping, see clause 5.
Because of the high saturation of core ( increasing of the flux density ) during the switching
impulse test, after each test impulse, to bring the transformer core to the normal beginning
condition (demagnetized), a few impulse tests which have small magnitude and positive
polarity, are applied to provide the duration needed for the next impulse voltage .
The switching impulse test is carried out on each line terminal of a three-phase winding in
sequence. During the application, neutral terminal is grounded, the windings which are not
under test are left open (grounded from one point). This connection type is like the one in the
induced overvoltage withstand test. The voltage distribution on the winding is linear like in the
induced overvoltage withstand test and the voltage magnitudes of the windings, which are not
under test, are induced according to the turn ratio. It must be noticed that at this instance the
voltage between the phases will be 1,5 times of the phase-to neutral voltage.
The test connections in three-phase transformers vary with the core design ( three of five-
legged), the limit of the voltage between the phases, the position of the delta connected
windings whether it is open or closed.
A B C
Rs
N
Cs
Rp
osc
osc
a b c V
R
Figure 16.1: Test circuit for the switching impulse test
ALSTOM 442/D
Test is first performed with a decreased voltage which is 50 % of the test voltage, then it is
carried out with full value impulse voltages the number of which is stated in the standards. The
peak value of the voltage is measured by an impulse voltmeter. The change of voltage
waveform and winding current are determined by means of an oscilloscope. The faults which
may occur during the test are determined by comparison of current and voltage ocillographic
records. The sudden collapses (flashover) in the voltage and abnormal sound effects show the
damage in the insulation of the transformer. The variation in the voltage waveform and the
increase of sound due magnetic saturation of the core must not be considered as reasons for
any fault.
Front time : T1 ≥ 20 ...< 250 µS = 1,67 T
Ninety – percent time : Td ≥ 200 µS
Time to first voltage zero : T2 ≥ 500 µS
Td
Td
0, 3
T
T2
Figure 16.2: Waveform of switching impulse voltage
ALSTOM 442/D
17. MEASUREMENT OF ACOUSTIC SOUND LEVEL
The purpose of the sound level measurement is to check that, the sound level of the transformer
meets the specification requirements, or quaranteed values given by the transformer
manufacturer.
The principal sound sources in transformers are :

a) core sound : caused by magnetostriction and inter-laminar magnetic forces.
b) load sound : caused by electromagnetic forces in the windings, tank wall, and
magnetic shields.
c) cooling equipment sound : caused by fans and pumps.
The predominant source of transformer sound is the core. The core sound depends on the flux
density in the laminations and the magnetic properties of the core steel. The low-frequency and
tonal nature. It occurs at twice the power frequency. Magnetic forces within the core will create
vibration and sound.
Load sound is the sound emitted by a loaded transformer in addition to its no-load sound. It is
caused by electromagnetic forces resulting from leakage fields. The load sound is proportional
to the fourth power of the current. The sources of this sound are the vibration of the tank walls,
magnetic shields, and the windings.
The core and winding sound dominates the intermediate freq. range between 100 and 600 Hz.
Sound produced by the cooling fans ( aerodynamic and motor/bearing sound ) is usually broad
band in nature. Factors that affect the total fan sound output, include type speed, blade design,
number of fans, and the arrengement of radiators. Pump sound is normally not significant if the
fans are running, although low frequency sound may be present.
The definitions have been given in IEC 60076 – 10.
The transformer must be located at the test site so that the free distance from the transformer to
reflecting objects is sufficiently large. It has to stand on a floor directly on its rollers, so that no
vibration will occur. All the accessories must be fixed tightly on the transformer.
The feeding of the transformer must be in sinus form at rated voltage and frequency.
When the tank height is less than 2.5 m the microphone is located at half of the tank height.
When the height of the tank is greater than 2.5 m the microphone position in the vertical
direction shall be at 1/3 and 2/3 of the transformer tank height. The cooling equipments
only energized, the microphon shall be on a horizontal plane at half of the height for cooler
structures with an overall height of less than 4m and for cooler structures with an overall height
of more than 4m the microphon positions shall be used which are horizontal plane at 1/3 and
2/3 of the height. The measurement is carried out alone all the circumference of the
transformer. The max. distance between the measuring points will be not more than 1 m.
The distance of the microphone from the principal radiation surface will be selected as follows;
1) If the fans of the cooling unit are switched-off, or they are assemble 3 m away from the
transformer, then the microphone must be 0.3 m away from the principal radiation
surface.
ALSTOM 442/D
2) If the cooling unit is switched-on ( the pumps and the fans working ) the microphone
must be 2 m away from the principal radiation surface.
The test object shall be energized as follows ;
a) transformer energized, cooling equipment and any oil circulating pumps out of service.
b) transformer energized, cooling equipment and any oil circulating pumps in service.
c) transformer energized, cooling equipment out of service, oil circulating pumps in
service.
d) transformer unenergized, cooling equipment and any oil circulating pumps in service.
Before starting the transformer sound level measurement the background noise level is
measured. If the difference between background noise level and the noise level of the
transformer is more than 8 dB (A), no correction at the noise level of the transformer is
needed. If the difference is between 3 dB (A) and 8 dB (A), a correction is needed acc. to the
standards. No measurement shall be made, if the difference between background noise level
and the transformer noise level is less than 3 dB. The correction factor for the effect of
background noise level to the transformer noise level is the in the below table as per IEC
60076 - 10;
The difference between the background The difference between the background
and transformer noise level. noise levels before and after the tests.
LpA0 − the higher LbgA initial LbgA − final LbgA decisions
≥ 8 dB - Accept test
< 8 dB < 3 dB Accept test
< 8 dB > 3 dB Repeat test
< 3 dB - Repeat test
The corrected average A- weighted sound pressure level, L , shall be calculated by using
pA
equation as follows :
L : The lower of calculated average background noise
bgA
= 10 log 10 pA0 - 10 bgA  - K
0,1L 0,1L
L L : Measured average sound level
pA0
pA  
K : Enviromental correction factor
The characteristics of the measuring equipment are described at the international standards.
The noise level of the transformer is measured acc. to IEC 60076 - 10, NEMA standards using
the evaluation of the weighting curve A.
The principle of parameters influencing noise are either internal ( frequency, flux, mass, quality
of magnetic material and operation ) or external ( distance ) . According to the laws of
acoustics, the volume of sound decreases with the square of the distance “ d “ from the
assumed point source, ie the centre of the equivalent hemispher ;
Lp(d ) = Lp(2m ) − 20. Log ( d /2 ) where ; d is in metres.
Other things the sound volume varies with the square of the frequency ;
Lp( f ) = Lp(50 Hz) + 20. Log ( f /50 Hz ) where ; f is test frequency.
ALSTOM 442/D
Figure 17.1: The position of the location of microphones at noise level test
ALSTOM 442/D
ROTATING MACHINES
Machinery 1
Generator: S = 5000 kVA synchronous

U= 6000 V
I = 833 A
f = 16 2/3 ÷ 50 Hz
Motor : P= 800 kW direct current
Generator: S= 2000 kVA synchronous

U= 6000 V
I = 192 A
f = 100 ÷ 200 Hz
Machinery 2
Generator: S= 3000 kVA

U= 6300 V
I = 275 A
2
f = 16 - 50-60 Hz
3
Motor : P = 1250 kW Asynchronous
TEST TRANSFORMERS
Transformer 1 S = 6000 / 6000 / 1500 kVA

U = 6300 V / 700........ 40000 V / 1000 V
I = 550 A / 660........87 A / 866 A
Transformer 2 S = 4000 kVA

U = 6300 V /700........40.000 V
I = 367 A /660.......58 A

U = 34500 ( 40000 ) V /77000 ( 88000 ) V
I = 335 A /150 A

U = 700 ..... 38000 / 6300 V
Transformer 5 S = 5000 kVA Booster transformer

U = 6000 ± 10 % V
ALSTOM 442/D
COMPENSATION CAPACITOR BANKS
BANK 1 : Group of capacitors with the following values.

- 13 groups, each of them with the rated voltage of 6,3 kV, totally 15,6 MVAr
- 2 groups, each of them with the rated power 6x100= 600 kVAr and the rated
voltage of 6,3 kV, totally 1,2 MVAr.
Bank 1 Rated power : 16.800 kVAr

Continuos operation : 20.300 kVAr
Power of 10 hours : 23.800 kVAr
BANK 2 : 216 pc. capacitors, each of them with the rated power of 200 kVAr, and the rated
voltage of 7,32 kV they can be connected to the system either with one phase or
three phases (∆ or )

BANK 3 : 180 pc. capacitors, each of them with the rated power of 300 kVAr, and the rated
voltage of 7,32 kV they can be connected to the system like Bank 2, either with
one phase or three phases (∆ or ).

MEASUREMENT TRANSFORMERS
Precision current transformers
3 pcs current transformers : 5-10-25-50-100-250-500-1000-1500 A/5A

30 VA, Class 0,1, 30 kV, 50/60 Hz
Manufacturer: Messwandlerbau, Bamberg
3 pcs current transformers : 25-50-75-100-250-500-750-1000-1250-1500 A/ 5A

15 VA Class 0,05, 30 kV, 50/60 Hz
3 pcs current transformers : 5-10-20 A/5 A

30 VA Class 0,2, 73 kV, 50/60/200 Hz
3 pcs current transformers : 100-250-500-800-1000 A/5 A

30 VA, Class 0,1, 73 kV, 50/60/200 Hz
Manufacturer: Ritz
3 pcs current transformers : 5 A/5 A

15 VA, Class 0,2, 45 kV, 200 Hz
Manufacturer: Ritz
ALSTOM 442/D
Precision voltage transformers
3 pcs voltage transformers : 3000 V/100 V

30 VA, Class 0,1, 50/60 Hz
Manufacturer: AEG
3 pcs voltage transformers : 10000 V/100 V

30 VA, Class 0,1, 50/60 Hz
Manufacturer: AEG
3 pcs voltage transformers: 30000 V/100 v

30 VA, Class 0,1, 50/60 Hz
Manufacturer: AEG
3 pcs voltage transformers: 3000 V/ 100 V

15 VA, Class 0,05, 50/60 Hz
Manufacturer: Messwandlerbau

15 VA, Class 0,05, 50/60 Hz

15 VA, Class 0,05, 50/60 Hz
3 pcs voltage transformers: 60000/ 3 V/100/ 3 V

30 VA, Class 0,2, 50/60/200 Hz
Manufacturer: Ritz
3 pcs voltage transformers: 66000/ 3 /100/ 3 V Isolation Cl. 73 kV

30 VA, Class 0,1, 50/60/200 Hz
Manufacturer: Ritz
ALTERNATING VOLTAGE TESTING EQUIPMENT’S
High-voltage test transformer: U= 350 kV/0,4 kV, 1- phase

S= 100 kVA I= 0,3 A/ 250 A f= 50 Hz
Manufacturer: AEG
High-voltage test transformer: U= 600 kV/ 3 kV 1-phase

S= 600 kVA I= 1 A/200 A f= 50 Hz
Manufacturer: AEG
Capacitive voltage divider : 350 kV/0,1 kV 50÷200 Hz

Manufacturer: AEG
Capasitive voltage divider : 800 kV/ 0,1 kV, 50÷200 Hz

Capacitance: 100 pF
Manufacturer: Haefely
ALSTOM 442/D
Voltmeter 2 pcs : U / 2 − U eff Voltmeter Digital

Class 0,5, 50 ÷200 Hz
Manufacturer: Haefely
IMPULSE VOLTAGE TESTING EQUIPMENT’S
Impulse voltage generator: No of stage :n = 10
Max. stage voltage :U L = 200 kV
Max. total voltage :U= 2000 kV
Max. total power :W= 200 kJ
Capacitance per stage : C= 1 µF
Manufacturer : Passoni+Villa
Multiple chopping device : 8- stage
Capacitance : 6000 pF/ stage
Lightning impulse voltage : ±1800 kV
Voltage divider : R-damped capacitive voltage divider
Switching impulse voltage : ±1450 kV
Capacitance : 400 ÷1600 pF
Ohmic voltage divider : Lightning impulse voltage : ±300 kV
Loading capacitor : Capacitance : 5 nF
Oscilloscope : Two channel
Type : 721
Accuracy : ±1%
Manufacturer : Haefely
Impulse voltage voltmeter: Accuracy : ±1%
Voltage : ±1600 V peak
Digital Imp. Meas.System : Type : TR-AS 200-12
Manufacturer : Dr. Strauss
Impulse Calib. System : KAL 1000 + Software IEC 1083-1 and IEC 60-2
KAL 1000- RIG (Unit step vol. Gen.) IEC 60-2
Manufacturer : Dr. Strauss
MEASURING BRIDGES
Schering measuring bridge : Type : 2801
Accuracy : 0,5%
Capacitance : 0...... 105 µF tan δ : 0...... 350%
Manufacturer : TETTEX
ALSTOM 442/D
Press gas condensator : Capacitance : 50 pF

Voltage : 400 kV a.c.
Ratio measuring bridge : Accuracy : ±0,1%

Voltage : 220 V a.c.
Measuring range : 1.........1000
Manufacturer : Hartmann+Braun
MEASURING INSTRUMENTS
Resistance measuring equip. : Transformer test system
Type : 2281
Accuracy : 0,1%
Measuring range : 10−6 .......102 Ω
Transformer test system
Type : 2285 c
Accuracy : 0,06 %
Measuring range : 10−6.......5 x 102 Ω
Type : 2292
Accuracy : 0,05%
Measuring range : 0,002….20000 Ω
Ratio measuring instrument : Transformer test system

Type : 2791 and 2 pcs 2793 + 2794
Accuracy : 0,1%
Measuring range : 0,18.........2000
Resistance measuring : Digital Low resistance ohmmeters

instrument Type : DLRO
Accuracy : 0,25%
Measuring range : 0,2 mΩ.......20Ω
Manufacturer : BIDDLE INSTRUMENTS
Digital Thermometer : 20 Channels, programmable

Type : DR 130
Manufacturer : YOKOGAWA
Power measuring unit : Wide Band Power Analyzer

(Wattmeter-Voltmeter- Type : D 6133 T
Ampermeter) 2 pcs Accuracy : 0,1%
Manufacturer : NORMA
Voltmeter-mean value : Digital, Interface

Accuracy : 0,1%
Manufacturer : NORMA
ALSTOM 442/D
Voltmeter Type : 3478 A Digital, Interface

Accuracy : 0,1%
Manufacturer : Hewlett Packard
Partial discharge measuring : Frequency : 30 kHz ÷ 2 MHz

equipment Measuring range : 0,1 pC.......50.000 pC
Type : 9120
Radio Noise Meter : Measuring range : 0,1 µV - 0,1 V

Type : EM- 7535
Manufacturer : Electro - Metrics
Insulation resistance meas. : Voltage : 5000 V. d.c.

instruments Measuring range : 500.000 MΩ
Type : SH2
Manufacturer : MEGGER
Voltage : 5000 V. d.c.

Measuring range : 1.000.000 MΩ
Type : BM 25
Manufacturer : MEGGER
Wave analyzer : Measuring range : Wave Analyzer 15 Hz......50 kHz

Type : 3581 A
Manufacturer : HEWLETT PACKARD
Noise measuring equipment : Meas. instrument : type 2033

1/3-1/3 octave filter: type 1625
Microphone : type 4145
Recorder : type 2317
Calibrator : type 4230
Manufacturer : Brüel & Kjaer
Vibration meas. equipment : Meas. range : 1.....1000ms−2 ,100 Hz.......10 kHz

Type : 2513
Calibrator : 4294
Manufacturer : BRÜEL & KJAER
Loss-factor meas. equipment : Meas. range : 0 ÷ 12 kV

Type : M2H
Manufacturer : DOBLE Engineering Company
Oscilloscope : 6- Channel
Type : 11401
Manufacturer : TEKTRONIX
: 2- Channel
Type : TDS 340
ALSTOM 442/D

The Testing of Transformer

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Testing of Transformer

Uploaded by

Copyright:

Available Formats

1

ALSTOM Elektrik Endüstrisi A.Ş.

Barış Mah. E 5 Altı, 1801 Sok. No 104

Tel. : + 90 262 648 33 00 Fax : + 90 262 641 20 35

Summary of the tests and measurements processes are given as follows:

1. Measurement of voltage ratio and check of vector relationship 3

TYPE TESTS AND SPECIAL TESTS

10. Temperature rise test 23

List of tests and measuring equipment of the test field 41

Prepared by : Haluk ODOGLU December 2001 ( 4. Edition )

1. MEASUREMENT OF VOLTAGE AND CHECK OF VECTOR RELATIONSHIP

Turn ratio measurement is performed by two separate methods;

c - Transformer to be measured d - The transformer with adjustable taps

Figure 1-1: Measurement of voltage ratio measuring bridge

the voltage of HV winding

Example : Vector group: Dyn5 B

Figure 1-3: Connection symbols for three-phase transform

2. MEASUREMENT OF WINDING RESISTANCE

Winding resistances can be measured by any of the following methods, by current-voltage

Figure 2-1: Winding resistance measurement by Current-Voltage method

The resistance measured by Thomson (Kelvin) Bridge method.

3. MEASUREMENT OF IMPEDANCE VOLTAGE AND LOAD LOSSES

1- Synchronous generator 5- Power analyser

Figure 3-1: Circuit for the load-loss measurement

E (%) = Eδ (%) + Ei (%) + Eu (%) Eδ ( % ) : measurement error

Ei ( % ) : current transformer turn ratio error

Since, phase displacement errors in measurement transformers are given in minutes.

DC losses = at tm measuring temperature Pdc = 1,5 ( I12 . RHV + I22 . RLV )

Additional losses = at tm measuring temperature Pac = Pkm - Pdc

The load losses at reference temperature:

4. MEASUREMENT OF NO-LOAD LOSSES

No-Load Losses and No-Load Current

1- Synchronous generator 4- Voltage transformers

Figure 4-1: Circuit for the no-load measurement

Category Highest voltage Tests

Uniform Um ≤ 72,5 type not applicable not applicable routine routine

Not 2 : If the ACSD test is specified, the SI test is not required.

The details of the partial discharge are given in section 8.

Repeated dielectric tests:

6. SEPARATE-SOURCE VOLTAGE WITHSTAND TEST

1- Transformer with adjustable voltages 5- Capacitive voltage divider

Figure 6-1: Test circuit for separate-source voltage withstands test

The current should remain constant during the test.

The test is successful if no collapse of the test voltage occurs.

7. INDUCED OVERVOLTAGE WITHSTAND TEST

a) Uniformly insulated windings

Transformers with Um < 72,5 kV normally , no partial discharge measurements

U 2 = 1,3 ⋅ U m / 3 phase – earth and U2 = 1,3 . Um phase - phase

Figure 7.2: Test duration voltage – time curve

b) non-uniformly insulated windings :

For three phase transformers, two test sets are required ;

On single –phase transformers, only a phase-to-earth test is required.

Long duration induced AC voltage test ( ACLD ) :

Uniformly and non uniformly insulated windings.

The voltages to earth shall be ; U1 = 1,7 ⋅ U m / 3 and U 2 = 1,5 ⋅ U m / 3

The partial discharge measurements shall be performed at all HV terminals.

8. PARTIAL DISCHARGE MEASUREMENT

The followings are determined during the measurement of partial discharges;

- To determine the existence of a definite partial discharge in the test object at a

- To determine the magnitude of the partial discharge at a predetermined voltage.

Creation of partial discharge and measured magnitudes

U: applied voltage Z: supply circuit impedance

Figure 8.1 a) An insulation having gas-filled cavity b) equivalent circuit